opengl - When does the transition from clip space to screen coordinates happen? -

I was studying the rendering pipeline and when I got the clipping phase, it was explained that this From the scene or in the camera) space, we have to go through the clip space , which is also called normalized device location (NDC), which is a cubic space for 1 to 1.

However, now I do not understand that the place is from place to place when it is from the screen:

1. Right to rafting After and after the clipping?

2. After rasterization and before scissors and Z-test?

3. Frame buffer at the end before writing?

Is one step away from NDC, to make all coordinates clip y NDC as a result of the NDC space [ -1 , 1 ] Anything outside matches a point which is outside the clipping volume. The reason for the coordination position before the NDC is called the clip space;)

Strictly speaking, ha However, the NDC space is not necessarily solid. It is true that the NDC space is a cube in OpenGL, but it is not in Direct 3D. Jade coordination in DDC space by 0.0 1.0 , while it behaves similarly between -1.0 to 1.0 GL in X and YGL and D3 (i.e. , They range from -1.0 to 1.0 ). NDC is a standard coordination spot, Feeling is the representation of different API.

Finally, screen NDC space (aka window space to space) during Rastraijeshn and define your viewport and depth range. Piece places will not really be understood in any other coordinate space, and this creates rasterization: pieces .

---

update:

The traditional OpenGL behavior is:

  glClipControl (GL_LOWER_LEFT, GL_NEGATIVE_ONE_TO_ONE);  

can be obtained through direct 3D behavior:

  glClipControl (GL_UPPER_LEFT, GL_ZERO_TO_ONE) ; // Inverted Y-axis to D3d